With growing energy demands and a need to switch to a sustainable source of energy key stakeholders are considering the use of high altitude wind energy systems. TU Delft and its start up company kitepower are key stakeholders investigating the commercial viability of this technology. With this goal the research group has developed several kite systems capable of accessing high altitude winds. It is believed that the low investment cost and high performance of kites could lead to a lower cost of energy. Concepts currently being considered involve a leading edge inflatable (LEI) kite that is controlled by an on-board control unit and is connected to a ground-based generator. Once the kite is deployed to the required altitude it enters a power generation stage where it is flown in a figure eight routine. This routine is controlled by the on-board control unit that pulls on tethers that are connected to the tip of the kite. This process is followed by a retraction phase where the kite is pulled back in. The goal of the system is to maximize energy production in the generation phase while limiting the energy consumed in the retraction phase. It is critical to assess and improve the kite design and its performance at all stages of operation such that the net power production can be maximized. While significant advancements have been made into the performance for normal flight there is a lack of research on the aerodynamic performance when there is control input to initiate a turn. The shape of the kite and the high angle of attack at which the kite is flown results in complex flow behavior involving separation, flow vortices, flow reattachment etc. This poses several challenges to maintain accuracy. A computational approach involving a steady-state Reynolds-averaged Navier Stoke (RANS) simulation is believed to be a computationally viable mode of analysis to capture the flow behavior. This thesis details the approach used to improve results attained using this method and understand the influence of deformations associated with control inputs on the aerodynamic performance of the kite. A control input is simulated using a finite element model (FEM) with the Abaqus software by reducing the length of the right steering and increasing the length of the left steering line by 0.5 m. This results in the right side reducing its curvature and being pulled towards the kite control unit (KCU). Whilst on the left side the bridle lines are less tensed, leading to an increase in curvature. The turning performance is governed by the offset and variations in magnitudes of forces. Meshes are generated that attempts to minimize geometry alterations whilst still maintaining high quality. The influence of boundary layer parameters are investigated. A trade-off is made where the influence of key parameters on accuracy and computational time is evaluated; where applicable improvements are made. Both the global as well as the local parameters of the kite in normal as well as turning orientation are analyzed. The results show that control induced deformations lead to a percentage reduction in the lift, whilst the effect on the drag is minimal. It is further seen that the kite initiates stall at an angle of approximately 40 degrees. The stall behavior is initiated at the mid-span of the kite and gradually moves to the tip. The turning performance is measured by looking at the yaw moment. The magnitude of this parameter is linked to the offset and magnitude in force vectors at the tip. Due to the delayed stall at the tips, it is observed that the yaw moment increases beyond 40 degrees. Accuracy issues using this method were seen when performing a validation study on a profile similar to that of the kite. These issues could be due to limitations of the method or potential errors in the reference study. It is recommended to reevaluate the validation study before using this method for detailed flow analysis. It is concluded, that in order to fully trust the relevance of the results, one would have to have to conduct a validation study. However the method’s ability to address non-linear flow effects within a limited time frame makes it a viable option for design optimization/system modelling.